WEAR MECHANISMS OCCURRING
IN PLASTIC GEARS

B. GOFFIN, R. LEGRAS and D. DEBIER
CERTECH (CEntre de Ressources TEchnologiques en Chimie) - UCL
Zone Industrielle C, B-7180 Seneffe, BELGIUM
e-mail: benedicte.goffin@certech.be

   The surface and bulk of injection moulded spur gears made of glass fibres reinforced polyamide internally lubricated with polytetrafluoroethylene PTFE have been characterised before and after wear testing on a gear test rig. The use of irradiated PTFE proved to be essential to obtain the best wear behaviour. The evolution of surface morphology across testing time allowed to visualise the formation and breaking of a film layer between the two running gears. This film, formed through the melting of the polymers at the surface, plays an important role in polymer tribology in reducing the dry wear. ln long lasting tests, the film is not only observed around the pitch line, but everywhere at the worn tooth surface. Generated wear debris could be parts of the film, broken through high shear and abrasive effect of fibre fragments, as confirmed by thermogravimetric and elemental analysis. The conclusions could be extended to various polyamide based composite gears. Running dissimilar materials against each other showed that the film is formed from the bulk of the gear. Therefore the film formation does not involve material transfer from one surface to its counterpart.

Key words: polymer gear, wear mechanism, transfer film, dry wear behaviour, polymer composites.

1. Introduction

         Plastic gears provide unique advantages over gears made of metal. Injection mouldable thermoplastic composites are being used increasingly in gear and bearing applications (Friedrich, 1993; Czichos et al., 1995; Reinicke et al., 1998). Dry wear resistance is one of the main reason for choosing polymer composites. Indeed in addition to a polymer matrix and a reinforcement, composites intended for tribological applications contain an internallubricating component which renders it possible to eliminate the need for external lubrication (Pratt, 1977).
         A huge arnount of polymer tribological data is available. However, this data is mostly based on plastic running against meta! in a pure sliding configuration (Plastics Design Library, 1995; Zhang, 1998; Friedrich et al., 1995). For plastic-on-plastic gear pairs, the wear behaviour is difficult to predict (Williarns and Quinn, 1995). Relatively little research has been carried out to study the wear mechanisms of polymer composites operating against themselves in non-conformal rolling-sliding contact (Chen et al., 1996). This work aims to understand the key morphological pararneters involved in the wear behaviour of thermoplastic composites gears.

2. Experimental

         Dry wear tests were conducted on a rig running spur gears against each other at a speed of 1000 rpmand a torque of 10 Nm.Indeed, it is important in a wear test to ensure representative service conditions (Hutchings, 1992; Zambelli and Vincent, 1998; Neale and Gee, 2000). These testing conditions correspond to a maximum sliding speed of 1 m/s and a contact stress of 46 MPa .
         The morphological characterisation of the worn surface of composite gears has been made by environmental scanning electron microscopy (ESEM) using a detector for backscattered electrons (BSE) and X-ray dispersive elemental analysis (EDX). The microscope is a Philips XL30 ESEM equipped with a field electron gun (FEG). For the observations, no sample preparation is required.
         Differential scanning calorimetry was performed on a DSC 821e from Mettler Toledo. Thermogravimetric analysis (TGA) was performed on a TGA 850 supplied r Mettler Toledo.
         Since polyamide (PA) and polyacetal (POM) account for 85% of gear materials used (Tsukamoto, 1995), gears based on polyarnide 6.6 matrix (PA 6.6) were chosen as model compound. The PA matrix was reinforced with 30% glass fibre (GF). Pol. ,. luoroethylene (PTFE) was used as internallubricant (15%).

3. Results

3.1. Effect of the PTFE type

         The type of internaI lubricant was examined. Figure 1 shows the observed morphology of two compounds having the same composition (PA 6.6 with 15% PTFE and 30 % GF) made from two different types of PTFE.
         The PA matrix appears in black, while the PTFE corresponds to grey regions as confirmed by EDX analysis. Pellets were cut transversally. Therefore the fibres appear as bright spots perpendicular to the plane of the picture.

Fig.1. Morphology of PA 6.6 pellets containing 30% OF and 15% PTFE (250x).

         In the case of non-treated PTFE, the dispersion is heterogeneous and large chunks of PTFE are observed (Fig.la). Using irradiated PTFE leads to a fine and homogeneous dispersion of PTFE in the PA matrix (Fig.l b ). The type of PTFE appears thus as an important parameter. The use of irradiated PTFE proved to be essential to obtain the best wear behaviour.
         Indeed, the adhesion between the matrix and the film formed at the wom surface (Fig.2) is higher when irradiated PTFE is used.

Fig.2. Film formed at the wom surface (200x),

3.2. Influence of the test duration

         We characterised gear surface and bulk before and after running in order to determine the major wear mechanisms. Gears made of the model compound {with irradiated PTFE) were tested against themselves during 24,48 and 72 hours. Different worn areas were examined, from the tip of the tooth, to the pitch line and the root. Figure 3 shows the morphology of a worn tooth surface after 24 hours of testing.
         After 24 hours, a film covers an important surface of the tooth around the pitch line. At the tip and near the root, this film seems less developed and is oriented due to wear as indicated by the white arrow in Fig.3.

Fig.3. Morphology of worn tooth surface after 24 hours of testing (200x).

         An estimation of the temperature generated near the surface was calcl11ated using the low melting peak of PA 6.6 observed by differential scanning calorimetry (DSC). Indeed thê melting behaviour of polyarnide is closely linked to the thermal history experienced by the material sample (Quintanilla et al., 1994). The results indicate a temperature of about 220°C. Nevertheless this temperature corresponds to a sample thickness of about 100 JLm. It is not the temperature at the extreme surface which is usually called the flash temperature. Furthermore, the DSC method is limited by the melting point of P A6.6. As a melted polymer film is observed on the wom surface, it is likely that the flash temperature is above the melting point of PA 6.6
( = 260° C ).
         The EDX analysis shows that the film is enriched in fluorine when compared to the unwom surface. However, it is not composed of PTFE alone. Therefore the PA matrix and the PTFE both contribute to the film formation. The film is thus probably formed by surface melting due to high temperature and high shear conditions encountered at the interface between the gears when the teeth are in contact.
         Near the root, fibres appear completely broken already after 24 hours. This could be explained by the reciprocating movement which occurs near the root of the tooth. This mechanism is more damaging than the rol1/slide wear.
         After 72 hours, the morphology is different from the one observed after only 24 hours of testing (Fig.4). The wear orient~tion is still observed. The film is formed everywhere at the worn tooth surface (Fig.4 a, b and c ), together with broken fibres.

Fig.4. Morphology of wom tooth surface after 72 hours of testing (200x)

         After 72 hours, a transition occurs on the wear curves and the wear starts to increase. The darnage is more severe and wear debris are generated. The TGA thermograrn of the wear debris (Fig.5) shows that PA 6.6 is degraded (a broad weight loss occurs earlier). Furthermore, the weight loss due to PTFE is not observed any more indicating that the content of PTFE is low. It is likely that PTFE is degraded. Therefore the wear debris could be parts of the film, broken through high shear and abrasive effect of fibre fragments, as confirmed by elemental analysis.

Fig.5. Thermogravimetric analysis (TGA) of a wom tooth surface after 24 and 72 hours, and the corresponding wear debris.

3.3. Varions polyamide based composite gears

         The formation of a film at the surface through the melting of polymers was also observed using polyamide 4.6, polyamide 6.10 and polyphthalamide as matrix material. As previously concluded with the model compound based on a PA 6.6 matrix, both polyamide and PTFE contribute to the formation of the film.

3.4. Dissimilar mating polymers

         It is well known that the primary wear mechanism for thermoplastics is adbesive wear. Adhesive wear (or interfacial wear) occurs when the counterface is smooth and is characterised by tbe transfer of polymer to the harder counterface (e.g. steel) (Hutchings, 1992). The tribological properties of polymers closely relate to this transfer film formation. ln this work, we showed that a film layer is formed between the contacting polymer surfaces. ln order to detect any polymer transfer, the wear behaviour of the model compound against unreinforced polyacetal (or polyoxymethylene POM) was studied.
         As previously observed, a film is formed at the wom tooth surface through melting of polymers. However, the interesting result is that the composition of the film at the POM tooth surface is close to pure POM. No fluorine could be detected by EDX analysis at the POM surface. At the nylon composite tooth surface, the film is made of melted PA and PTFE. Therefore, the film is formed from the bulk, and its formation does not involve material transfer from one surface to its counterpart.

4. Conclusions

         A morphological approach was followed in order to understand wear mechanisms occurring in plastic gears. Complementary research (e.g. thermal and elemental analysis) was performed to link the morphological observations with the wear behaviour .
         A film is created during nylon based composite gear run through the melting of P A and PTFE at the surface of the tooth. This film plays an important role in the tribology of polymers in reducing the dry wear .The evolution of surface morphology across testing time helped us to visualise the formation and breaking of the film layer between the two mnning gears. Running dissimilar materials against each other (PA 6.6 composite against POM) showed that the film is formed from the bulk of the gear. Thus in the case of polymer gears mnning against each other, the wear mechanism does not imply transfer of material from the surface to its counterpart.

Acknowledgment

         The authors wish to express their thanks to the European Union for financial support through Brite-Euram funding (contract no. BRPR -CT98 - 0703). Furthermore, the help and assistance provided by the Davall Moulded Gear Company in supplying gears, and the School of Manufacturing and Mechanical Engineering (University of Birmingham) in testing gears c are gratefullyacknowledged.

References